: Patients treated with electroconvulsive therapy typically present with the most severe forms of major depression. Likely due to the increasing representation of medication-resistant patients, ECT response rates have diminished relative to earlier decades. This diminished response rate and early relapse following response are critical clinical problems in the use of ECT. Using the CSMD mechanism, this project addresses two key issues in the optimization of ECT in patients with major depression: whether patients treated with ECT should receive concurrent treatment with antidepressant medications (to enhance ECT outcome and/or prevent early relapse) and the role of electrode placement (high dosage right unilateral (RUL) ECT versus low dosage bilateral (BL) ECT) in maximizing short-term response and minimizing side effects. Patient enrollment, treatment, and evaluation will be conducted at Wake Forest University, Washington University, and the Western Psychiatric Institute and Clinic, with staff at the New York State Psychiatric Institute responsible for study coordination and monitoring. The study uses a random assignment, double-masked, parallel group design with two phases. In Phase 1, stratified by the classification of medication resistance, patients are randomized to concurrent treatment with nortriptyline (NT, n=210], venlafaxine (VEN, n=210) or placebo (PL, n=210), and simultaneously to high dosage (6 times threshold) RUL ECT (n=315) or low dosage (1.5 times threshold) BL ECT (n=315). Based on substantial preliminary data, the hypotheses will be tested that (1) compared to PL, concurrent NT or VEN results in superior symptomatic response, without a meaningful difference in side effects, and (2) RUL and BL ECT are equal in efficacy, but with significant advantages to high dosage RUL ECT in the magnitude of short- and long-term cognitive side effects. Support for these hypotheses in a large and diverse sample should have widespread ramifications for clinical practice. In the Phase 2 double-masked, 6-month continuation trial, remitters who received PL during ECT are randomized to NT and lithium (LI) or to VEN-LI. Patients who had been randomized to concurrent NT or VEN during ECT receive continuation treatment with NT-LI or VEN-LI, respectively. Standard practice involves the discontinuation of antidepressant medications prior to ECT, the abrupt discontinuation of ECT upon response, and then a switch to continuation pharmacotherapy. This practice likely diminishes response to ECT and heightens relapse in the first several weeks following ECT. Phase 2 of this study, centering on the comparison of patients treated with an antidepressant medication (NT or VEN) or placebo during ECT, will provide the very first data on whether starting an antidepressant medication from the beginning of the ECT course reduces the rate of early relapse.